 When we think of science, we often think of the scientific method. But calling it the scientific method is a bit misleading. There's no single scientific method. There are a number of ways that scientists can improve their understanding of how the natural world works. And it doesn't happen in a single step. The process of building scientific understanding happens over time. Let's explore the different processes that contribute to our scientific understanding. One approach is called the deductive method. This involves first coming up with a hypothesis, a possible explanation of how the world works, developing a better understanding of the mechanisms of the natural world. In science, a hypothesis is no good unless it can generate predictions. Then, scientists collect observations to see whether that prediction comes true. For example, in 1859, John Tyndall made some predictions about what warming from increasing greenhouse gases should look like. He predicted that if greenhouse gases were causing warming, then we should see nights warming faster than days and winters warming faster than summits. Around 150 years after that prediction, scientists finally had the opportunity to test Tyndall's prediction. It turns out his prediction was right. The data confirmed his prediction and is one piece of evidence for greenhouse gas warming. Let's look at a counterexample where the data disproves a hypothesis. In the early 20th century, some scientists thought that the oceans were big enough to soak up all the extra carbon dioxide we were emitting from fossil fuel burning. In the 1950s, Charles Keeling began taking detailed comprehensive measurements of atmospheric carbon dioxide. Year after year, he took the most accurate comprehensive measurements of CO2 that had ever been taken. Within a few years, it became clear that atmospheric carbon dioxide was increasing. The oceans weren't absorbing all of our CO2 emissions. Some of it were staying in the atmosphere. The data showed that the notion that the oceans would take up all our CO2 emissions was incorrect. Another scientific approach is called the inductive method. This involves collecting the data first, then scientists analyze that data to see if they can detect any patterns. Once a scientist has analyzed the data and developed a hypothesis, whether it be using the inductive or the deductive method, the job isn't done yet. Before research can get published in a scientific journal, it gets scrutinized by other scientific experts. This process is known as peer review. What it strives to do is weed out errors, making sure the research is rigorous and evidence-based. Peer review is not a guarantee that the research is perfect. Mistakes still get published in scientific journals. But because of the level of scrutiny, I would argue that peer reviewed research is the highest quality source of scientific information available. Okay, so a scientist has collected data, analyzed it, developed a hypothesis, survived peer review and got their research published in a scientific journal. The job's done, right? Not even close. When scientific research has been published, the next step is other scientists check the results by trying to replicate it. They run their own experiments or take their own measurements to see if they obtain the same result. When a result is replicated independently, we have more confidence that it's accurate. When a result is replicated by completely different types of measurements, then our confidence is even stronger. For example, there are a number of different ways to check whether humans are causing global warming. Observing nights warming faster than days or winters warming faster than summers are just two human fingerprints. Our climate is rich and complicated and we expect to see a whole range of different human fingerprints. For example, we should see the upper atmosphere cool at the same time that the lower atmosphere warms. It's a completely independent observation to changes in the daily or yearly cycle, but the conclusion is the same. Humans are causing global warming. When independent lines of evidence all point to the same consistent conclusion, scientists call this a conciliance of evidence. As the body of evidence builds up, what follows is overwhelming agreement among the scientists, a scientific consensus. In and of itself, a consensus doesn't prove that a hypothesis is correct. Our confidence in a scientific proposition comes from the evidence. Scientific consensus is a reflection of our scientific understanding. The common thread running through all of this is a reliance on scientific evidence. Whether the scientific method is inductive or deductive, evidence is a crucial part of the process. Our confidence in the result strengthens as more lines of evidence build up. It should be only when a conciliance of evidence has developed that a scientific consensus forms.